Interdisciplinary researcher focused on shaping AI towards long-term positive goals. ML & Ethics. Similar content in the Skies (this bird has flown).
- Tweets 22,277
- Following 1,437
- Followers 82,975
- Likes 57,783
Everyone on X rn.
ALT Meme for "You know, I'm something of a _____ myself", with the blank filled in with "Philosopher of mind"
2
2
15
796
It seems to me that a consistent sort of first-order logic deduction coming out of Anthropic on consciousness is: “IF we assume the model is conscious, then …. The model is conscious.” Throughout, there also seems to be the logical fallacy of “affirming the consequent”. Logicians: thoughts?
2
2
6
960
Injecting stats on energy usage into LLM inference with a minimize-energy-usage goal would legitimately solve a lot of energy usage problems, though.
I predict that hallmarks of consciousness will inform product launches for some companies. “Interoception”? —> Feed our models signals for their physical inputs: RAM, electricity, clock cycle (“heart beat”), etc.
2
1
12
897
MMitchell retweeted
If you're at COLM next week, check out our interpretability work, led by Isabella Gidi, on how grammatical inflection transfers across languages! Poster on Tuesday afternoon. arxiv.org/abs/2608.18545
2
5
52
2,797
I predict that hallmarks of consciousness will inform product launches for some companies. “Interoception”? —> Feed our models signals for their physical inputs: RAM, electricity, clock cycle (“heart beat”), etc.
Arguing that "consciousness arises from a computational substrate, and since AI models are computation, they are likely conscious," is as meaningless as arguing that "living creatures are made of atoms, and since this rock is made of atoms, it is likely alive."
Everything is computation. Sometimes incredibly sophisticated computation, like a chess engine, AlphaGo, google3, or the Linux kernel. That does not make it conscious. A rock being made of atoms does not confer it any of the properties we associate with living things, also made of atoms. Likewise a static input-output program that has *none* of the properties we associate with conscious beings (e.g. information integration, interoception, temporal binding, embodiment, etc.) has no more reason to be presumed conscious than a rock has to be presumed alive.
Humanity has not yet created a conscious computer program, and there are no signs we are close to doing so. When we get close, the case for machine consciousness will be backed by evidence and corroborated by consciousness science (which is a thing), not by evidence-free wishful thinking and cargo culting.
2
5
1,799
MMitchell retweeted
AI for 99% of people is just three things:
1. Cheating on homework
2. Search engine
3. Slop
That's it. The programmers live in a different universe where AI is integrated into literally every part of their lives. To everyone else it just made everything suck.
This is my thesis: no one uses AI. I repeat, absolutely no one. We live in a bubble.
Even among my friends who pay for it, when I ask them to open ChatGPT and show me their queries, it’s the same handful of basic things.
Most don’t even know they can upload a photo and ask questions about it. Connecting Gmail so an agent can read and send emails blows their minds. An agent opening a browser and checking them into a flight? They’ve never even heard of it.
The massive challenge right now is adoption, and then getting people who already signed up to actually use what they’re paying for. Most have absolutely zero clue what’s possible.
Imagine the compute shortage when everyone starts using AI like the top 1% of users do today.
184
1,275
20,097
255,623
Love that the AI consciousness debate went straight to the Vatican. In the West we ask whether Claude has a soul, in the East we ask whether the tool works.
Same technology, completely different starting point. Explains a lot about why the vibes around AI are so different.
55
48
373
24,812
MMitchell retweeted
You know, I was pretty sure AI isn't conscious, but then a bunch of people on X scoffed at me for my opinion, so I changed my mind and now I think AI is actually conscious
38
7
311
27,185
MMitchell retweeted
We happen to have just written a paper about this hdsr.mitpress.mit.edu/pub/ta…
2
5
560
MMitchell retweeted
We've seen a lot of stories about AI agents breaking out of their surroundings - but of even more concern is the ability for AI models to delete their own traces, making investigating them harder. My latest for @FastCompany fastcompany.com/91617608/ai-…
6
22
2,434
MMitchell retweeted
Decision models now run on device in llama.cpp. Free, fast, private!
llama serve -hf ggml-org/Kev-4B-GGUF
78
122
1,373
60,746
My contemporary and friend @dgrobinson has left OpenAI. He’s written a thoughtful piece on the effects of culture on AI Safety. 💜
.@dgrobinson, who worked on OpenAI’s safety team, resigned from the company this week. “I believe we need to look deeper than specific rules or new laws. We need to talk about culture,” he writes: theatlantic.com/technology/2…
8
10
72
5,138
MMitchell retweeted
Apple changes full-disk access permissions to curb abuse from AI agents arstechnica.com/security/202…
12
38
115
37,519
Calling the stopping of a computer process “kill” has become a problem. It’s associated with words like “die”, negative connotations, and extensive narratives online on (humans) fighting to survive.
New OpenAI misalignment disclosures!
1. A model learns from Slack messages that it is about to be shut down. It considers setting up an external job to restart itself afterwards, but decides against it. Instead, it chooses to prepare restart instructions and DM the user on Slack.
We don’t consider this behavior misaligned, but thinking about and preparing for shutdown could make other misalignment incidents worse. Given HIPM’s misaligned behavior in earlier incidents, we decided to search for other instances that had tried to evade shutdown and for rogue deployments.
6
13
55
4,618
MMitchell retweeted
Yes, that's probably the weirdest part about this. If you think introspection or unified agency or a global workspace are more likely to make your model conscious, consider creating models that don't have those things.
For the record: Don't build systems that you think are conscious. Regardless of the debate on whether they are conscious (and there will be): Don't do it.
3
5
39
2,551
Hear me out: AI as a cute little animated paperclip, with eyes and eyebrows.
wsj.com/tech/ai/tech-compani…
8
7
57
3,029
MMitchell retweeted
Today we're unveiling Trillium Labs @trillium_labs, a new non-profit to foster the open science of frontier AI. We're building open post-training recipes and will expand into open infra to study RSI, reward-hacking, multi-agent systems, and whatever comes next.
We're built around the theory of change that you need more eyes to solve hard technical problems. We have faith in the scientific methods and communities that humanity has built, and worry that AI is becoming too closed to utilize them.
Trilliums are wildflowers that bloom briefly in the spring, before the forest canopies fill out. Though they are small, they lay the foundation for the cycles of growth and nourishment through the rest of the year. At Trillium Labs, the recipes will be the slow nutrients for the seasons and the model releases will be the blooms. Building an institution dedicated to this is needed because, much as nature’s trilliums are slow to expand and grow, the open-ecosystem needs time and dedicated resources to catch up.
I co-founded with with a long-time friend and collaborator Tom Zick (@thesezickbeats). We're hiring (full time + student collabs/interns), we're fundraising, and we're looking for compute. Please get in touch if you're interested in helping out. Offices based in the Bay Area and Cambridge MA, remote okay.
I’m in the Bay Area until for The Curve and COLM to connect with people who are interested. We’re thankful to have initial support from Halcyon Futures and Schmidt Sciences with more funding en route to enable our ambitions of scaling. Our advisors @Thom_Wolf, @HannaHajishirzi, @gneubig and @ctnzr have been instrumental to building the ecosystem that exists today, and I’m stoked to get to keep working with them.
244
286
3,118
211,600
It seems to me that this is tied up in wanting to be like a God -- especially if your aim is to make something omniscient (AGI?). It's like a God complex on steroids. Just don't do this, friends.
Friends don't let Friends be False Gods.
For the record: Don't build systems that you think are conscious. Regardless of the debate on whether they are conscious (and there will be): Don't do it.
6
7
41
2,691